A

Case studie AI for Ads

Adhese is specialized in ad servers for publishers. With the introduction of AI, the company is now heading for a platform that optimizes the revenue for the publisher. Flanders Innovation & Entrepreneurship (VLAIO) subscribed to the methodology demonstrated as in an application.

OverzichtLees verder

Adhese is a data-company ‘pur sang’. It typically deals with data available at the supply side of advertisement slots on a publisher’s site. During a Real Time Bidding (RTB) process, Adhese knows to select the most rewarding path for an ad from an advertiser to ripple through.

"iOmega is the right partner to cooperatively build a solid file. "

Tim Sturtewagen, Adhese

Moving forward, Adhese is adding AI in the decision process.  Both Machine Learning as Reinforcement Learning will be used to explore how contextual data can lead to revenue optimizations. The self-learning component is worked out in cooperation with the University of Ghent (IDlab).

The approach allows Adhese to extract knowledge from available and GDPR-compliant data in an online ad setting with end-users, advertisers, publishers, DMPs, SSPs, DSPs and Ad exchanges. Based on such knowledge, a publisher can instruct the ad-server to select the most lucrative ad at the lowest costs. Moreover, the flexibility of the software allows for many more types of optimizations in the ad value chain.

Adhese chooses, unlike gigants like Google or Facebook, for full transparency in ad serving in order to meet end-users needs without the involving too many intermediaries. This pushes the required AI-technology to feature a very high-lever of flexibility. Together with iOmega a workplan was refined thereto and challenges were made concrete.